\begin{table}

\caption{Long-Run Effects on 2022 Registration and Voting\label{longrun}}
\centering
\begin{tabular}[t]{lccc}
\toprule
  & Baseline Registration & (Re)Registration & Turnout\\
\midrule
Information-Only Mail (T1) & \num{0.0042} & \num{0.0047} & \num{0.0032}\\
 & (\num{0.0043}) & (\num{0.0049}) & (\num{0.0035})\\
Research-Informed Mail (T2) & \num{0.0077} & \num{0.0066} & \num{0.0044}\\
 & (\num{0.0043}) & (\num{0.0050}) & (\num{0.0036})\\
\midrule
Difference: T2-T1 & 0.0035 & 0.0019 & 0.0012\\
 & (0.0045) & (0.0052) & (0.0038)\\
Num.Obs. & \num{23768} & \num{23768} & \num{23768}\\
Control Mean & 0.3692 & 0.1101 & 0.0558\\
\bottomrule
\multicolumn{4}{l}{\rule{0pt}{1em}All models include pre-treatment registration status as a covariate}\\
\multicolumn{4}{l}{\rule{0pt}{1em}* p $<$ 0.05}\\
\end{tabular}
\end{table}
